CharMark: character-level Markov modeling for interpretable linguistic biomarkers of cognitive decline
Bibliographic record
Abstract
Dementia, one of the most prevalent neurodegenerative diseases, affects millions worldwide. Understanding linguistic markers of dementia is crucial for elucidating how cognitive decline manifests in speech patterns. Current non-invasive assessments like the Montreal Cognitive Assessment (MoCA) and Saint Louis University Mental Status (SLUMS) tests rely on manual interpretation and often lack detailed linguistic insight. This paper introduces a first-of-its-kind interpretable artificial intelligence (IAI) framework, CharMark , which leverages first-order Markov Chain models to characterize language production at the character level. By computing steady-state probabilities of character transitions in speech transcripts from individuals with dementia and healthy controls, we uncover distinctive character-usage patterns. The space character “ ”, representing pauses, (treated here as the space token between words rather than acoustic pauses), and letters such as “n” and “i” showed statistically significant differences between groups. Principal Component Analysis (PCA) revealed natural clustering aligned with cognitive status, while Kolmogorov-Smirnov tests confirmed distributional shifts. A Lasso Logistic Regression model further demonstrated that these character-level features possess strong discriminative potential. Our primary contribution is the identification and characterization of candidate linguistic biomarkers of cognitive decline; features that are both interpretable and easily computable. These findings highlight the potential of character-level modeling as a lightweight, scalable strategy for early-stage dementia screening, particularly in settings where more complex or audio-dependent models may be impractical.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".